Saturday, 13 February 2016

Revenge of the Noobs

6 February 2016 - Charlton Athletic 0 Bristol City 1

One of the debates that’s been taking place since the appointment of Lee Johnson, which took place an hour or so before kick-off, is the experienced vs inexperienced manager question; do we want a big-name who has a better chance of getting us promoted, or do we want an up-and-coming manager as a gamble which might pay off?

What’s interesting about that debate isn’t so much the question itself, which has been done to death now, as the very basis of the discussion. Whichever camp a person is in, they tend to accept the basic premise – that a name manager would have a greater chance of taking the club on in the short term, potentially as far as the top six or even the Premier League. The division seems to be between those who think that’s something to aim for in and of itself, and those who think we should be cannier and strategic, even if it carries the risk of missing an opportunity to progress. Even the club’s official channels have been complicit in this thinking, with Johnson asked in his first interview whether the club had taken a ‘risk’ bringing him in – not a question you’d expect Nigel Pearson, say, to have been posed.

But I’d not seen any quantification of this received wisdom; any demonstration that a manager with 200 Premier League games under his belt is automatically a better short-term choice than a man with 200 games in League One. Sure, there’s an apparent logic in the idea that someone who’s worked with higher-calibre players can teach ours more, that someone who’s achieved success knows how to do so and can import that knowledge. And there are the examples of a safe pair of hands taking an underachieving club into the top flight – Neil Warnock at QPR, say, or Steve Bruce at Hull. But how much of that is a sign of a genuine trend, and how much is confirmation bias? I wanted to find out.
So I’ve looked at the data. Specifically, I looked at every managerial change made in the Championship between the 2005-06 season and the 20014-15 one, inclusive – 144 managerial changes made by 42 clubs, from Barnsley to Wolves. That’s a pretty clear set of data.

Then, the slightly trickier bit. Defining precisely who is a ‘Name’ manager and who is ‘up-and-coming’ is difficult. Defining success or failure is tough, too, particularly given that most managerial appointments end up in failure. Nevertheless, I had a good go. I worked on the principle that a Name manager had managed in a major top flight and/or internationally, while an Up-and-coming manager hadn’t managed at an equivalent level to the Championship before. This left a middle category, the Experienced manager, someone who’d had at least one job in the Championship or at a comparable European league before being hired. Inevitably this led to judgement calls – is the Scottish Premiership comparable to the Championship? You could argue that it is, but having managed St Johnstone meant that Derek McInnes felt like an Up-and-coming rather than Experienced manager. But the three categories broadly seemed to work. The majority of appointments – 60 – were unsurprisingly of Experienced managers, with Up-and-coming coaches receiving 49 jobs and Name managers being tempted into the remaining 35.

(Sometimes the same manager can be considered in multiple ways, of course, as their career progresses – Phil Brown is an Experienced manager when he takes the Hull job, but that job makes him a Name as he moves into the Premier League, so he is in that category when Preston bring him in. And Lee Clark’s tenure at Birmingham gives him the Championship experience to move him from Up-and-coming to counting as Experienced by the time Blackpool come in.)

Quantifying success and failure was even harder, and I realised very early that there had to be a middle category here – Neither. How many managers took over a club around 18th in the middle of the season, took them to 8th the following year, but then got sacked with the club 19th in season three? It would be arbitrary to describe either of these as success or failure, so I didn’t. Broadly, promotion, a top-six finish in the final season, or improvement on improvement before departure counted as Success – relegation, being in the bottom three when sacked, or never achieving good quality results throughout a tenure counted as Failure.

There are more Failures than anything, perhaps unsurprisingly – 60 out of those 144 appointments ended badly. Nearly as many ended in a great big 'meh' – 56 of 144. Only 28 appointments, less than 20% of the total, ended in what feels to me like a real Success.

A lot of this, of course, is attempting to quantify opinions, so in order to show my working I've created a Google Doc you can check and use as a basis to argue with me.

(A side note on our team. You’ll notice that every Bristol City manager appointed at this level, from Coppell onward, qualifies as a Failure under my system, and that spans a Name in Coppell, two Up-and-coming managers in Millen and McInnes, and the Experienced O’Driscoll. Indeed, Bristol City are the only team in the entire list to have hired as many as four managers, all of whom have to be judged as failures. It's fairly clear that the common link in their performance is the club, rather than the ability or otherwise of a set who've collectively won promotions, managed and coached in the Premier League, and threatened to break the Celtic stranglehold on the SPL. More than anything, the fact that no other Championship club however poorly run have been so hard to succeed with speaks to the state the club had got itself into toward the end of the Gary Johnson era, and the amount of work most of those managers were obliged to do to try to change things around.)

Anyway. With the facts, such as they were, at my disposal, the analysis was the easy bit. Let's look first at those 28 successes.

42% of managers were Experienced, and 43% of Successes were achieved by Experienced managers – all things being equal, exactly what you'd expect. Tony Pulis gets Stoke to the Premier League. Nigel Pearson does the same for Leicester. Paul Hart, of all people, takes a short-term contract to save Crystal Palace, and does so. Hiring an experienced manager can clearly work.

More interesting results came from looking at the other two. 34% of hires were Up-and-coming, yet they accounted for 39% of successes – really overperforming, thanks to, say, Owen Coyle at Burnley, Brian McDermott at Leicester, or (right now) Gary Rowett at Birmingham. And those 24% of managers with Names? They accounted for just 19% of successes, making the likes of Bruce and Warnock much more like the exceptions than the rules.

Given the existence of the 'Neither' category, it wasn't a sure thing that the failures would follow the same rule. But while those 24% of Names had only been 19% of the successes, they made up a full 30% of the failures, thanks to the likes of Ian Holloway at Millwall, Malky Mackay at Wigan, and Steve McClaren at Forest. The Up-and-coming managers weren't terrific at avoiding failure – 34% of hires, 32% of failures (thanks Andy Thorn, Uwe Rosler and Jim Gannon) – but it's still a little less than the law of averages would predict, and a far better performance than the Names. Experienced hires were just about the safest bets, at 38% failure to 42% of hires, but that's not a notably more significant over-performance than the Up-and-comers.

There we have it, then. The idea that hiring a big-name manager is a guaranteed route to the land of ambrosia and nectar is a nonsensical one, not borne out by statistics at all. And while a wily head at this level has the slightest of slight chances of doing better than a bright young thing, there's very little in it, and certainly not enough to justify a preference on general principle for one over the other.
Because this is about general principles. None of this means that Lee Johnson is guaranteed to outperform his relegation rival at Rotherham, Neil Warnock. But what it does mean is that Warnock having managed in the Premier League compared to Johnson's third tier experience is not a reason to prefer one to the other. Indeed, based on experience alone you'd predict that Johnson has a better chance than Warnock. He may not do as well, the same way that in 2006 an up-and-coming crop of Parkinson, Wise, Grant, Simpson and Waddock were outperformed by Mick McCarthy. But then, who won the league that season? Tyro boss Roy Keane's Sunderland, that's who.

And if bringing in an up-and-coming manager isn't a risk (and it clearly, now, isn't), it might just have a higher level of reward attached. Look at the nine Premier League clubs we've played as equals over the last decade or so, now achieving far more than us: Leicester, Southampton, Watford, Stoke, Crystal Palace, West Brom, Bournemouth, Swansea and Norwich. Only one, Palace, were taken up by a Name manager, Ian Holloway, who built on the good work done by the up-and-coming Dougie Freedman and was then sacked to make way for someone better suited to the Premier League. Sacking the manager who got you up is pretty much the way of things, and only three of the clubs haven't done it – the last three. Those managers? Eddie Howe, Brendan Rodgers and Alex Neil.

I had to stick to my system when classifying Howe and Rodgers, whose spells at other Championship clubs made them both technically Experienced, but neither had the sort of CV on their initial appointment at Bournemouth and Swansea that City fans would have been excited about. In terms of a club going from the bottom division to being a sustainable challenger at the top, these two clubs are the big stories, and they've both achieved the final stage under the guidance of a manager who is, at best, a pretty inexperienced Experienced higher.

Both are run in an intelligent, far-sighted way; both are cited as models for a club like City to attend to; and both had the gumption to believe that a manager with no record of doing anything like it could take them into the Premier League.

If we want to be anything like those clubs, we be excited about doing what they did – giving a chance to someone with everything to prove, and then allowing them the best possible shot at doing so.